Journal of Medicinal Chemistry, I 9 73, Vol. 16, No. 5
574
Notes
Table I. Bulbring and Wajda Anesthesia Test
2 Control (vehicle) Control (procaine) P-PEG400' P-TEG~
Number of squeals out of 5 at time (min) 5 15 30 45 60 75 90
5
5
5
5
5
5
5
0
0
0
2
3
4
5
0 0
0 0
0 0
0
0
0
0
2 0
3 0
105
5
5
4 3
5 4
~~
'Procaine attached to polyethylene glycol 400. bProcaine attached to tetraethylene glycol. Table 11. Rabbit Eye Corneal Anesthesia Grade at mina
5 10 15 30 45 60 Procaine 0 1 2 2 2 2 Vehicle 2 2 2 2 2 2 P-PEG400 1 0 0 1 2 2 P-TEG 0 0 0 0 1 2 %radation ofblink response: 0 is no blink, i.e., anesthesia; 1, sluggish response; 2, blink, i.e., no anesthesia.
such as ether, benzene, and chloroform and were insoluble in water. The procaine derivatives were subjected to two local anesthesia tests. The first one was according to Bulbring and Wadja,"in which guinea pigs had their backs shaved and 1% solutions of the substances and a control of vehicle and procaine (as a base) were used. The substances were first dissolved in 1 drop of DMF and made up to volume with sterile distilled water. Intracutaneous injections were made of 0.1-ml volumes to give a weal. Immediately afterward every 5 min a needle was jabbed into the weals five times in succession. The response of a squeal and skin contraction indicate that the animal feels the pain, while absence means local anesthesia. Five areas and three guinea pigs were used per group (Table I>. It can be seen that the effect of procaine lasts in this test between 30 and 45 min, having no effect at 75 min. P-PEG 400 lasts about twice as long and has no activity at 105 min. P-TEG shows even better prolongation of action. The second test was rabbit eye corneal anesthesia. It determines lipid soiubility or penetration effects. In this procaine is weak. Two rabbits were used per group, and 0.5 ml of 1% solutions of substances was instilled into the conjunctional sac of their eyes and held bathing the eye surfaces for 30 sec. Every 5 or 10 min thereafter the eyes were touched with a fine stylus. This normally causes a blink. If there is local anesthesia the eye remains open on being touched with the stylus (Table 11). In this test P-PEG 400 shows a slower onset lasting for 20 min which is twice as good as procaine. P-TEG shows an immediate onset and lasts for 45 min. This result parallels the previous result and shows that P-TEG is a promising prolonged action local anesthetic. The preliminary pharmacological results indicate that the attachment of procaine to polyethylene glycols has increased the duration of activity of the drug. However, the fact that the shorter polymer showed the longer duration of action may point out that increase of duration of activity is not only a matter of the molecular weight but also of other factors such as different permeabilities through biological membranes, partition coefficients. etc. Experimental Section Materials. Tetraethylene glycol and polyethylene glycol 400 (number-average molecular weight = 400), Fluka, were used. Ir spectra
were carried out on a Perkin-Elmer 257 instrument and nmr on Varian T-60. w,w '-Dichlorocprbonate of Tetraethylene Glycol. Tetraethylene glycol (19.4 g, 0.1 mol) and a solution (240 ml) of phosgene in toluene (12.4%)was stirred for 48 hr at room temperature. Excess of phosgene and toluene were driven off in vacuo at low temperature, leaving the product: 28 g (88%);ir 1778 cm-' (OC(=O)Cl); nmr (CDC1,) 6 3.6 (m, 12, CH,OCH,), 4.4 (m, 4 , CH,OC(=O)Cl). Anal. (C,,H,,Cl,O,) C, H, C1. The w,w'-dichlorocarbonatc derivative of polyethylene glycol 400 was similarly prepared in 88%yield. Anal. (C,H,O),~,(C,Cl,O,) C, H, c1. iV,N'-(p-~-Diethylaminocarbethoxyphenyl)-w,w'dicarbamoyltetraethylene Glycol (P-TEG). A solution of procaine hydrochloride (2.73 g, 0.01 mol), triethylamine (6.06 g, 0.06 mol), and chloroform (35 ml) was cooled in an ice bath, and the dichlorocarbonate (1.6 g, 0.005 mol) was added dropwise during 30 min with stirring. The reaction mixture was stirred overnight. Dry ether was added, the precipitated triethylamine hydrochloride was filtered, and the filtrate was evaporated. The oily residue was dissolved in CHC1, and precipitated by petroleum ether. The procedure was repeated twice. The remaining oil was purified by preparative tlc on silica gel (Merck PF 254) using absolute ethanol as eluent: yield 2.0 g (56%);ir 850 (1,4-C6H4-), 1695 (C,H,COO-), 1710 (-NHCOO-), 1100 cm-' (-CH,OCH,-); nmr (CDCl,) 6 1.0 (t, 12, -CH,), 2.85 (q, 8, -NCH,), 6.9 (m, 8, -C6H4-), 3.5 ( s , 12, -CH,OCH,). A d . (C,6H,4N401,)C, H, N. The reaction product between w,w'-dichlorocarbonate of polyethylene glycol 400 and procaine hydrochloride was similarly prepared and purified: yield 65%;ir and nmr confirm the structure. Anal. (CH,CH,O),.,(C,,H,,N,O,) C, H, N.
Acknowledgment. We wish to thank Dr. H. Fenton of Pharmatox, Israel, for carrying out the pharmacological tests. References ( 1 ) C. E. Hall and 0. Hall, Experientia, 17, 544 (1961); 18, 38 (1962). (2) R. J. Cornel1 and L. G. Donaruma, J. Polym. Sci., Part A , 3, 827 (1965). (3) L. J. Donaruma and J. Razzano, J. Med. Chem., 9, 258 (1966). (4) T. C. Merigan and M. S. Finkelstein, Virology, 35, 363 (1968). (5) N. Minor, T. Alfrey, J. Koehler, and R. Zimmerman, Bacterial Proc., 225 (1971). (6) G. P. Lampson, A. K. Field, A. A. Tytell, M. M. Nemes, and H. R. Hilleman,Proc. SOC.Exp. Bzol. Med., 135, 911 (1970). (7) A. Einhorn, K. Fiedler, C. Ladisch, and E. Uhlfelder, Justus Liebigs Ann. Chem., 371, 142 (1909). (8) H. L. Smitz and A. S. Loevenhart, J. Pharmacol., 24, 159, 167 (1924). (9) W. Schulemann, Klin. Wochenschr., 3,676 (1924). (10) I. G. Farbenindustrie A.-G., German Patent 582,715 (1933). ( 1 1 ) R. Fussganger and 0. Schumann, Arch. Exp. Pathol. Pharmakol., 160, 53 (1931). (12) J. J. Bonica, Curr. Res. Anesth. Analg., 30, 1 , 76 (1951). (13) K. Soehring, K. Scriba, M. Frahm, and G . Zoellner, Arch. Int. Pharmacodyn. Ther., 87, 301 (1951). (14) C. G. Hunter, D. E. Stevenson, and P. L. Chambers, Food C o m e t . Toxicol, 5, 195 (1967). (15) H. F. Smyth, Jr., C. P. Carpenter, and C. S. Weil, J. Amer. Pharm. Ass., 44, 27 (1955). (16) H. F. Smyth, Jr., C. P. Carpenter, and C. B. Shatter, ibid., 36, 157 (1947). (17) E. Bulbring and 1. Wajda, J. Pharmacol., 85, 78 (1945).
Nitro Analogs of Pyoluteorins Colin G. Hughes and Alun H. Rees* Department of Chemistry, Trent University, Peterborough, Ontario, Canada. Received August 29, I 9 72
A requirement for antibiotic activity in 2-(2'-hydroxybenzoy1)pyrrole (la) is one or more electron-withdrawing groups X and Y.' Compound l b is the parent antibiotic pyoluteorin, a metabolite of Pseudomonas aeruginosa with antibacterial and antifungal activity.2 Replacement of the
Journal of Medicinal Chemistry, 1973, Vol. 16, No. 5
Notes
In our previous work on monodeoxypyoluteorins' we prepared 5-dechloromonodeoxypyoluteorin(1 h) via its methyl ether. It had low activity but we were able to utilize the ether l i by converting it to its 5-nitro derivative l j which was demethylated to the phenol l k which had improved activity (Table I).
1 a,R=R'=H b,X=Y=Cl;R =OH;R=H c, x = R = H ; Y = C l ; R ' = O H , d, X = N O , ; Y = H ; R = M e ; R = M e 0 e.X=NO,;Y=R =H;R=Me f, X = NO;;R = R = H; R' = OMe g, X = NO,; Y = R = R' = H h , X = R = R =H;Y=Cl 4 X = R ' = H ; Y=Cl;R=Me j, X = NO,; Y = C1; R' = H; R = Me k, X = N0,;Y = C1; R = R' = H
Discussion
5-chlorine atom by hydrogen gives 5-dechloropyoluteorin (1 c)',~of much lower activity. Because of the presence of a nitro group in the pyrrolnitrins (2),5-' which are also metabolites of Pseudomonas species, we decided to investigate the effect of a nitro group in place of chlorine on the activity of the pyoluteorins. We therefore nitrated 2-(2',6'-dimethoxybenzoy1)pyrrole' and 2-(2'-methoxybenzoyl)pyrrole' which gave the 5-nitro derivatives Id and l e , respectively. These were demethylated to 2-(2'-hydroxy-6'methoxybenzoyl)-5-nitropyrrole (1 f) and 2-(2'-hydroxybenzoyl)-5-nitropyrrole (lg). Preliminary attempts to further demethylate If were unsuccessful but in view of the generally greater activity' of monodeoxy compounds (Table I), we decided to concentrate on this series. However, from
H
0 2
3a
575
OMe
H-0
OMe
3b
one preparation of Id, an interesting by-product was obtained with mol wt 321 (mass spectrum), indicating the presence of two nitro groups. This was confirmed by the nmr spectrum which showed two AB systems (J3,4= 4.5, J 4 ' , 5 t = 9 Hz). The ir spectrum contained a weak carbonyl band and these facts are consistent with structure 3 in which the enol form 3b is significant. This is also indicated by the fact that the OH-NH peak is rather broad and the M-OH peak in the mass spectrum at m/e 304 is 12% which is higher than usual in the ethers. Also,in the nmr, there was no coupling between H(l) and H(3) or H(4) before or after exchange of H( 1) by deuterium, again indicating the predominance of the enol form. We have previously invoked a related enolization as the mechanism in our first synthesis of pyoluteorin (formula 11, see ref 9).
We have previously shown' that replacement of the chlorine of monodeoxypyoluteorin by more electropositive halogen, uiz. bromine and iodine, results in decreasing antibiotic activity. The accompanying lower acidity of the pyrrole NH group was thought to be the reason for this. A significant increase in the NH acidity, which could cause stronger binding to a receptor site, would be brought about by the substitution of a more strongly electronegative group such as nitro and should therefore be accompanied by improvement of microbiological activity. That this expectation was realized is shown by compound l g in which the chlorine atoms of monodeoxypyoluteorin are replaced by a 5-nitro group. It is further shown (Table I) that when only the 5-chlorine atom is replaced by nitro giving l k , the activity is further slightly enhanced, in line with expectation.
Experimental Section Infrared spectra were run on a Unicam SP 200,nmr spectra [TMS internal standard (7values quoted)] on a Jeol C60HL machine, and mass spectra on the AEI MS 12 at 70 eV on the probe. Microanalyses are within prescribed limits. Melting points are uncorrected. 2(2'-Methoxybenzoyl)-5-nitIopyrrole (le). Fuming nitric acid (0.12ml) was added to 0.4 g of 2-(2-methoxybenzoyl)pyrrole in glacial acetic acid and the mixture was kept at 70-80" for 3 hr. The solvent was evaporated and the residue, in chloroform solution, was washed with sodium bicarbonate solution, dried (MgSO,), and passed down a silica column, eluting with chloroform-petroleum ether (bp 30-60') (3:2),giving 0.36 g of nitro compound: 75%; long yellow fibers; mp 116.5-117.5' (crystallized from benzenepetroleum ether); mass spectrum m/e 276;re1 intensity (100,M'); nmr, see Table 11. Anal. (C,,H,$rl,O,) C, H, N. 2-(2'-Hydtoxybenzoyl)-5-nitropyrrole(lg). Demethylation of the above methyl ether l e with boron trichloride in methylene chloride8 gave the free nitrophenol (lg): mp 146-147' (crystallized from benzene) in 85% yield; mass spectrum m/e 232 (70,M'), 120 (100).Anal. (C,,&N,O,) C,H, N. Nitration of242',6 -Dimethoxybenzoyl)pymle. By the above method, 2-(2',6'dimethoxybenzoyl)pyrrolegave the 5 -nitro derivative Id: yellow cubes; mp 230-234" (crystallized from benzenepetroleum ether); mass spectrum m/e 276 (100,M'); nmr, see Table 11. Anal. (CI,H,,N,O,) C, H, N. Using acetic anhydride as solvent, a similar nitration, 3 hr a t 70-80",gave in addition to the 5-nitro compound above, a byproduct, separable on tlc (silica, ether-petroleum ether, 3:l) as a gum which could not be crystallized. It had only a medium intensity ir band at 1660 cm-I compared with other analogs. The compound was not very stable but on the basis of its spectra it was assigned the structure 2-(2',6'dimethoxy-3'-nitrobenzoyl)-5-nitropyrrole(3):
Table I. Comparison of Activitiefl of Pyoluteorin Nitro Analogs with Those of Deoxypyoluteorin8 Complete inhibition, pg/ml Medium Organism Incubation time 1g lk Deoxypyoluteorin If > 100 N. agar Staph. aureus 209-P 0.8 3.1 18 hr 1.5 >loo N. agar 0.8 3.1 Staph. aureus (resistant) 18 hr 1.5 100 N. agar Shig. flexneria 2a 18 hr 6.2 3.1 >loo Shig. flexneri (resistant) N. agar 6.2 >loo 100 18 hr 6.2 Gndida albicans >loo S. agar 6.2 25 >loo 2 days S.agar 6.2 100 Tricophyton asteroides 7 days 1.5 1.5 100 S . agar 12.5 7 days 1.5 6.2 Tricophyton interdigitale S. agar 6.2 100 7daw 0.8 1.5 i'?iCODhVtOnrubrum =Plate method, solvent acetone (maximum concentration lo%), nutrient agar medium; culture temperature 37 and 27",single determinations.
Table 11. Nmr Spectra of Methoxy- and Hydroxybenzoylnitrop yrroles‘ Compd
le b3 Id If 3
H(3) 3.47 3.35 3.75 3.67 3.6
H(4) 3.05 2.9 3.08 3.03 3.1
NH
OH
-0.53
OMe
Solvent
6.22
CDCl, DMSOd, DMSOd, DMSOd, CDCl,
Exchanged
-3.62 -3.62 -0.5
0.37
6.38 6.38 6.25
‘In compounds 1, J l , , = 2 Hz; in 3 H(1) is not split. mass spectrum m/e 321 (90,M+), 179 (100);nmr, see Table 11. S-Nitro42’-hydroxy-6’-methoxybenzoyl)pyrrole (If). Demethylation of the dimethyl ether Id by the method above gave 75% of monomethyl ether: mp 139-141” (crystallized from benzene-petroleum ether); mass spectrum m/e 262 (65,M’), 150 (100). Anal. (CI2H,,N,O,) C, H, N. When boron tribromide was used for demethylation, a minor product was also obtained, shown by mass spectrometry to be a monobromo derivative of If with the bromine in the phenolic ring. Reparation of 4Chloro-2-(2’-methoxybenzoyl)-5-nitropyrrole (lj). Nitration of 0.2 g of 4-chloro-2-(2’-methoxybenzoyl)pyrrole in acetic acid as above gave 80 mg of nitro derivative: m p 141-142“; yellow cubes from benzene-petroleum ether; mass spectrum m/e 280 (50, M’), 135 (100). Anal. (Cl2H&1N,O,) C, H, C1, N. 4-Chloro-2-(2’-hydroxybenzoyl)-5-nil1opyrrole (lk). Demethylation of lj as above gave the phenol lk: 65%; mp 183-185” (crystallized from benzene-petroleum ether); mass spectrum m/e 266 (80, M’), 120 (lOO).Anal. (Cl,H,CIN,O,) C , H, C1, N.
Acknowledgments. We thank the National Research Council of Canada for financial support and the Sankyo Co. of Tokyo for microbiological results. References (1)
Notes
Journal of Medicinal Chemistry, 1973, Vol. 16, No. 5
576
K.Bailey, G. R. Birchall, D. G. Durham, C. G. Hughes, and A. H. Rees,Int. Congr. Pure Appl. Chem., 23rd, Abstr., 73
(1971). (2) R. Takeda, Hakko Kogaku Zasshi, 36,281 (1958). (3) R. Takeda, Bull. Agr. G e m . Soc. Jap., 23, 126 (1959). (4) G.R. Birchall, C. G . Hughes, and A. H. Rees, Tetrahedron Lett., 4879 (1970). (5) H. Imanaka, M. Kousaka, G. Tamura, and K. Arima, J. Antibiot., Ser. A , 18, 205 (1965). (6) M.Hashimoto and K. Hattori, Chem. Pharm. Bull., 14, 1314 (1966). (7) M.Hashimoto and K. Hattori, Bull. Chem. SOC.Jap., 39,410 (1 966). (8) D. G.Durham, C. G. Hughes, and A. H. Rees, Can. J. C h e m , 50, 3223 (1972). (9) K.Bailey and A. H. Rees, ibid., 48,2258 (1970).
Hydroxylamine Derivatives as Poten tial Inhibitors of Nucleic Acid Synthesis? John B. Hynes,* Glen R. Gale, Loretta M. Atkins, David M. Cline, and Kenneth F. Hill Departments of Pharmaceutical Chemistry and Pharmacology, Medical University of South Carolinu, and Veterans Administration Hospital, Charleston, South Carolina 29401. Received December 14. I972
The antineoplastic activity of hydroxyurea l a has been the subject of numerous investigations.‘ Currently, its NH,CON(R)OH la, R = H b, R = CH, c, R = C , H , ?This work was supported by Grant CA-11811from the National Institutes of Health, by a South Carolina Appropriation for Research (J. B. H. and G. R. G.), and by Student Fellowships from NIH General Research Support Grant to the Medical University of South Carolina, 5 SO1 RR 05420 (D. M. C. and K. F. H.).
principal clinical utility lies in the treatment of chronic granulocytic leukemia.2 Its N-methyl (lb) and N-ethyl (IC) derivatives have also been studied in considerable detail3 but neither of these was shown to have a clear-cut chemotherapeutic advantage over the parent compound. The analog of la, hydroxyoxamide (oxamylhydroxamic acid) 2a, as well as its acetylated derivative acetoxyoxamide 2b, was shown to be a selective inhibitor of in vitro DNA synthesis having a similar level of potency to that of NH ,COCONHOR 2a, R = H b, R = COCH,
However, 2a displays considerably higher toxicity due at least in part to the fact that it readily oxidizes hemoglobin to methemoglobin4 Other close structural analogs of la such as 1- and 3-hydroxybiuret (3a and 3b) have also been investigated for antimitotic effects.6 NH,CON(OH)CONH, 3a
NH 2CONHCONHOH
3b
In the present investigation, two modes of modification of 2a were chosen in an effort to obtain reduced toxicity: (a) insertion of a substituent, preferably a bulky one, on the N’ nitrogen and (b) attaching a group at the hydroxyl which would be less susceptible to hydrolysis than 2b. Compounds 5-13 (Table I) are representative of the former approach while 1 4 and 15 are examples of the latter. Two synthetic routes were evaluated for the preparation of ”-substituted hydroxyoxamides. The more general route involving the reaction of ethyl N-hydroxyoxamate (4) with the appropriate amine was found to be unsatisfactory since complex mixtures were invariably formed even at low temperature. Therefore, these compounds were prepared from their corresponding esters, the properties of which are summarized in Table 11. Although admittedly equivocal, the structures of compounds 14 and 15 have been assigned the 0-carbamyl configuration. This is based upon the fact that neither compound gave a positive reaction with aqueous iron(II1) chloride. In addition, the infrared spectra of 14 and 15 have a band in common with 2a and 2b at 2.94-2.96 /.which i is absent in the spectra of 7 and 11. Three new pyrimidine-5-carbohydroxamicacids 16-18 were prepared as potential antimetabolites by the reaction of their corresponding esters with hydroxylamine in the presence of excess base. The analogous reaction employing 5-carbethoxycytosine, however, resulted in hydrolysis to the corresponding carboxylic acid. Each of the compounds presented in Table I was evaluated for inhibition of DNA, RNA, and protein synthesis in Ehrlich ascites tumor cells in vitro. Seven compounds showed inhibition of DNA synthesis at concentrations below 10-3M and these are presented in Table I11 together with the corresponding data for l a and 2a. It will be seen that none of the present compounds display the degree of selective inhibition of DNA synthesis which is shown by l a and 2a, although the activity of 14 is considered significant. With the exception of 8, none of the derivatives of 2a (5-7 and 9-1 5 ) were found to form methemoglobin (cfi EXperimental Section). However, the structural modifications generally also resulted in loss of selective action. Only compounds 7 and 9 produced “unbalanced growth” of Escherichia coli as evidenced by the formation of long, filamentous cell forms which is often associated with selectivity of action against DNA synthesis. It will be noted
